Abstract 1887 Substrate Recognition by Two 3' to 5' RNA Polymerases in Dictyostelium discoideum

Unlike most polymerases that act in the 5′ to 3′ direction, tRNAHis guanylyltransferase (Thg1) synthesizes RNA 3′ to 5′. Thg1 catalyzes an essential reaction during tRNAHis processing by adding a G nucleotide on the 5′ end of the tRNA in most eukaryotes, forming an identity element for aminoacylation of the tRNA. Recent characterization of Thg1 homologs with alternative specificities, known as Thg1-like proteins (TLPs), raises questions about how distinct substrates are recognized by different members of this highly conserved enzyme family. In the slime mold Dictyostelium discoideum (Ddi), DdiThg1 adds G to the 5′ end of cytosolic (cy-) tRNAHis, while DdiTLP2 catalyzes the same reaction, but only with mitochondrial (mt-) tRNAHis substrates. In vivo and in vitro, these two enzymes exhibit strict specificity for their respective tRNA substrates. Moreover, unlike Thg1, DdiTLP2 does not depend on the GUG anticodon for tRNAHis recognition, suggesting different mechanisms for tRNA recognition by these two enzymes. We aimed to determine the molecular basis for the distinct RNA substrate specificities of DdiThg1 and DdiTLP2, and thus to provide insight into the general mechanisms of RNA recognition utilized by distinct 3'-5' RNA polymerases. To answer this question, electrophoretic mobility shift assays (EMSA) were used to assess tRNA binding by DdiThg1 vs. DdiTLP2. These assays revealed no difference in either enzyme's ability to bind to different tRNAs, requiring additional catalytic factors to explain the selective in vitro activities of these enzymes. Sequence comparison was used to identify a unique residue in DdiTLP2 (R187) that is different from an absolutely conserved D/E residue at the analogous position in the rest of Thg1/TLP family members, including in DdiThg1. We hypothesized that this unusual R187 residue may be responsible for the unique biochemical properties exhibited by DdiTLP2. Indeed, replacement of the conserved D residue in DdiThg1 with the R found in DdiTLP2 caused a dramatic reversal of DdiThg1 substrate specificity. The DdiThg1 D150R variant lost the ability to catalyze G-1 addition with cy-tRNAHis and instead gained the ability to act on DdiTLP2's mt-tRNAHis substrate. Interestingly, kinetic and conservative amino acid replacement studies revealed that the change in DdiThg1 predominantly impacts the second step of the 3'-5' addition reaction, suggesting a direct role for the conserved D residue in this nucleotidyl transfer step. This result may also help to explain a previously observed in vivo growth defect associated with alanine replacement of the analogous D residue in Saccharomyces cerevisiae (Sc) Thg1, despite the fact that the ScThg1 D153A variant enzyme retains full in vitro catalytic activity. These results suggest that the DdiThg1 D153/ScThg1 D153 residue helps to control RNA substrate specificity and carry out the nucleotidyl transfer reaction step in this unusual family of 3'-5' RNA polymerases. Grace Johnecheck was supported by NIH T32 GM141955; this research was funded by R01 GM087543.